We evaluated relationships among two circulating molecular forms of brain natriuretic peptide (BNP32 and NT-proBNP), severity of hypertension (HTN), and cardiac hypertrophy in subjects with mild, moderate, and severe HTN. We prospectively studied 78 patients (43 males; mean age 51.4 +/- 11 yr) with essential HTN and 28 age- and sex-matched controls. BNP32 and NT-proBNP were measured by radioimmunoassay. In grade 1 HTN, BNP32 was not elevated and NT-proBNP was reduced (P = 0.030) compared with controls. However, log-transformed values of BNP32 and NT-proBNP were both increased with severity of HTN from grade 1 to 3 (P <0.0001 and P = 0.003, respectively). By multivariate analysis, log BNP32 was independently predicted by age (beta = 0.210, P = 0.026) and HTN grade (beta = 0.274, P = 0.004), whereas log NT-proBNP was independently predicted by sex (beta = 0.235, P = 0.012) and HTN grade (beta = 0.218, P = 0.0023). Two forms of BNP were measured in normal subjects and patients with essential HTN. In grade 1 HTN, BNP32 was unchanged and NT-proBNP was significantly reduced compared with controls. As severity increased in humans with grade 1 to 3 HTN, both BNP32 and NT-proBNP levels were increased while not being affected by the presence of left ventricular hypertrophy. The lack of activation of BNP32 together with the reduction of NT-proBNP in grade 1 HTN may represent an impaired response of the BNP system in the early phase of HTN. The later activation of both forms of BNP may be a late compensatory effect, because it correlates with severity of HTN rather than cardiac hypertrophy/remodeling.